Idiopathic pulmonary fibrosis therapy development: a clinical pharmacology perspective

Abstract

Drug development for idiopathic pulmonary fibrosis (IPF) has been challenging due to poorly understood disease etiology, unpredictable disease progression, highly heterogeneous patient populations, and a lack of robust pharmacodynamic biomarkers. Moreover, because lung biopsy is invasive and dangerous, making the extent of fibrosis as a direct longitudinal measurement of IPF disease progression unfeasible, most clinical trials studying IPF can only assess progression of fibrosis indirectly through surrogate measures. This review discusses current state-of-art practices, identifies knowledge gaps, and brainstorms development opportunities for preclinical to clinical translation, clinical populations, pharmacodynamic endpoints, and dose optimization strategies. This article highlights clinical pharmacology perspectives in leveraging real-world data as well as modeling and simulation, special population considerations, and patient-centric approaches for designing future studies.

Introduction

Idiopathic pulmonary fibrosis (IPF) is a subtype of interstitial lung disease (ILD) that has no known cause or cure. A poor understanding of disease etiology, combined with an unpredictable course of disease, results in poor patient outcomes comparable to those of patients with lung cancer. ¹ ILD is characterized by cellular proliferation, interstitial inflammation, fibrosis, or a combination of these within the alveolar wall that is not due to infection or cancer. Interstitial fibrosis is usually the predominant phenotype of ILD, with idiopathic interstitial pneumonia (IIP) diagnosed if no cause is identified. Notably, IPF is the most common IIP. ² IPF is a progressive disease that manifests as symptoms of increasing cough and dyspnea leading to loss of lung function and impaired quality of life with a median survival of 2–3 years in patients who do not receive treatment.^3,4 Currently, IPF prevalence is similar to that of many cancers, with worse survival prognosis and increasing incidence worldwide.^5–7 Diagnosis of IPF in clinical practice also varies from region to region, using combinations of high-resolution computed tomography (HRCT), lung biopsy, lung function tests, and process of elimination to rule out potential causes of ILD. ⁸ To date, while no curative treatments exist for patients with IPF, two anti-fibrotic drugs, pirfenidone and nintedanib, have been approved and have been shown to be effective in slowing disease progression and are currently considered the standard of care. ⁹ Indeed, while both pirfenidone and nintedanib reduce the rate of forced vital capacity (FVC) decline and pooled data and meta-analyses suggest potential benefit on reducing mortality, ¹⁰ neither of these anti-fibrotic drugs cures IPF patients by reversing the course of fibrosis and both drugs are associated with substantial tolerability issues. ¹¹

Undoubtedly, there is a clear unmet need to find a curative therapeutic for patients with IPF. Yet, the path to resolution is still hindered by challenges that span from early research to clinical success. For example, there is currently no standardized practice for preclinical dose extrapolation to first-in-human (FIH) trials, no established pharmacodynamic (PD) biomarker that can be used to establish a direct exposure–response relationship, and no method of directly measuring disease progression in patients with IPF. Nevertheless, many breakthroughs and novel strategies have emerged in the past decades, including efforts to better understand the mechanism of disease progression, development of novel preclinical models, and utilization of new strategies in clinical trial designs. ¹² Among these, one challenge that still stands is dose selection and optimization strategy throughout the drug development process. In this review, we aim to describe the current ‘best practices’ from preclinical to clinical development, challenges associated with these practices, and opportunities from a clinical pharmacology perspective.

Disease background and progression in animal models and in human patients

One of the key challenges in preclinical to clinical translation of IPF therapies is the lack of an animal model that recapitulates all the clinical and biological manifestations of human IPF. While a number of models have been developed and used in the preclinical space, each with unique features that has helped to facilitate better understanding of individual features and mechanisms of this complex disease, many gaps remain in their individual clinical translation capabilities. A thorough review of the various IPF animal models and their unique features has been published by Tashiro et al. ¹³

When leveraging animal models to support FIH study design, a few key factors should be taken into consideration: (1) the model’s acuity of disease progression at the time of treatment (inflammatory phase versus fibrotic phase), (2) treatment duration (animals cannot be treated for a long time due to natural reversal of disease), (3) species differences in target expression and engagement, and (4) identification of clinically relevant endpoints and biomarkers that will support dose and regimen optimization. The timing of intervention in animal models may be important in identifying the prophylactic or therapeutic effect of an investigational drug. For example, the classic bleomycin-induced IPF rodent model shows rapid disease manifestation, where the initial phase is representative of the inflammatory stage of human IPF disease, while the fibrotic stage of disease does not manifest until 14–21 days after initial induction.^14,15 Furthermore, the frequency of bleomycin dosing necessary to produce desired fibrogenesis has been reported to be strain-dependent, thus adding another layer of variability to study design and outcomes. ¹⁵ Thus, the nature of pulmonary damage induced by bleomycin in rodents compared to humans is also significantly different in pathogenesis and mechanisms of pulmonary injury. ¹⁵ Despite the imperfections of the bleomycin-induced murine model, it has been accepted as the basis of approval for both currently marketed drugs, nintedanib and pirfenidone, and is still widely used to support FIH studies.

Another consideration when using nonclinical models is the timing of intervention. In most clinical practices, patients are identified during the fibrotic stage of disease and rarely, if ever, during the inflammatory stage of IPF. Notably, the duration of treatment is also limited in some animal models. In this regard, bleomycin-induced rodents have been observed to have a natural partial reversal of disease over time as the ‘disease’ itself was chemically induced. ¹⁶ However, human fibrosis is a slowly progressing and chronic condition. Thus, lack of chronic disease in animal models limits our understanding of the long-term anti-fibrotic effects of an investigational drug.

Current clinical practice

Diagnosis

Clinical diagnosis of IPF also varies from region to region, based on the practicing physician’s experience and knowledge. However, in 2022, the ATS/ERS/JRS/ALAT published an updated clinical practice guideline to provide recommendations intended to provide the basis for rational, informed decisions by clinicians. Though there may be certain differences and nuances with other guidance documents published prior to 2022, such as the Fleischner Society White Paper, which has also been acknowledged internationally by scientific communities and subject experts as IPF diagnostic and treatment guidance documents, the similarities that both of these guidelines agree upon highlight the unanimous current understanding of IPF disease. ¹⁷ Due to the lack of more quantitative/definitive symptoms or disease biomarkers, and the highly heterogeneous nature of disease manifestation in IPF patients, the time to diagnosis is reported to be a median of more than 1 year from the onset of disease symptoms. ¹⁸ Taking all of these factors together, IPF patients who qualify for participation in clinical studies have vastly different baseline disease and unpredictable clinical outcomes. For example, having learned from previous studies (CAPACITY 004 and CAPACITY 006) that demonstrated conflicting results, the pirfenidone ASCEND phase III study revised the patient selection criteria to be more selective in patient pulmonary function and disease severity at screening. ¹⁹ Perhaps in an attempt to account for some of the variabilities in patient population, and due to a lack of a clear definitive and uniform diagnosis guidance, the investigators’ preferences for inclusion and exclusion criteria for patient enrollment also diverge in stringency and disease severity. A review of the completed and ongoing phase II and phase III IPF trials that are published on clinicaltrial.gov revealed that patients tend to be above the age of 40 years, with confirmed diagnosis of IPF ranging from 3 months to 5 years (though this does not equate to disease duration), use of HRCT scan as a part of inclusion criteria, and baseline lung function parameters that range from a minimum FVC of 40–55% range, predicted diffusing capacity of the lungs for carbon monoxide (DLCO%) range of 25–40% minimum, and a 6-min-walking distance (6MWD) range of 100–150 m. An overview of eligibility criteria of all ongoing phase II (n = 25) and phase III (n = 6) studies is summarized in Figure 1.

Figure 1.

Eligibility criteria for active and ongoing phase II/III trials since 2018.

Only ongoing phase II and III trials studying IPF that started on or after January 1, 2018, were included in this figure. Values are expressed as percentages. Trials with more than one endpoint were accounted for in multiple categories.

Endpoints

The FVC either in the form of absolute or derived values (% predicted FVC) is the most commonly employed and accepted clinical endpoint in IPF clinical trials by the health authorities.^20–24 One of the advantages of FVC as a clinical endpoint is its ease of measurement. In multiple studies, FVC at baseline and its change throughout the course of the study have been shown to be significant covariates and are associated with disease progression and survival. Small changes in FVC can result in significant clinical worsening. Patients with a % predicted FVC decrease in at least 10% to less than 15% had significantly increased risk of death, with a hazard ratio of 2.2; while patients with a decline of 15% or more had an even greater hazard ratio of 6.1, compared to patients with less than 5% decrease from baseline. ²⁵ The variability of FVC data is also well documented. ²⁶ Since FVC can be measured noninvasively and repetitively, pooled across studies, and has significant association with mortality in patients with IPF, it continues to be the gold standard for efficacy evaluation in drug development for IPF.

Another metric that is frequently evaluated in IPF clinical trials is the 6MWD. It is measured during the 6-min-walk test (6MWT), in which patients are asked to walk as far as possible over 6 min on a flat surface. During this test, supplemental oxygen dependency is also recorded and evaluated as a covariate. Patients with IPF usually have reduced exercise capacity, and 6MWD at baseline and the change of 6MWD over 24 weeks were shown to be independent predictors of mortality in patients with IPF. ²⁷ Notably, the pooling of 6MWD data across trials for analyses requires careful consideration to ensure the 6MWT was conducted consistently across study sites and trials (e.g. standard scripts for words of encouragement, facilities that can accommodate straight line walking, etc.). The results of a 6MWT are affected by numerous factors, including age, body size, comorbidities, and the use of supplemental oxygen during the test, and these need to be borne in mind when interpreting the results of individual and serial tests. ²⁸ To aid in standardizing the procedure for this outcome measure, in 2002, an ATS statement was published to provide practical guidance for the 6MWT. Specifically, it reviewed detailed factors that influence results, such as those mentioned above, presented a brief step-by-step protocol and outlined safety measures. ²⁹ Building on this, in 2014, the ERS/ATS technical standard was developed by a multidisciplinary and international group of clinicians and researchers with expertise in the application of field walking tests, including the 6MWT. ³⁰ This technical standard for field walking tests reflected the best available evidence regarding procedures that should be used to achieve robust results.

Acute exacerbation of IPF and respiratory-related hospitalizations are also clinical endpoints with profound implications. Patients with IPF have significantly increased risk of mortality if experiencing an acute exacerbation or respiratory-related hospitalizations.^25,26 Unlike FVC and 6MWD, acute exacerbation of IPF and respiratory-related hospitalization data are not longitudinal data but can be analyzed as a time-to-event (TTE) endpoint and correlated to survival prediction. Studies have shown that patients with IPF who experienced acute exacerbation usually have a median survival of 3–4 months.^31,32 Other common endpoints in IPF clinical trials include patients-reported outcome/quality-of-life questionnaires and composite endpoints, such as progression-free survival (PFS) and TTE. Out of 36 ongoing phase II and phase III IPF studies registered on clinicaltrials.gov that started on or after January 1, 2018, 8 studies listed composite endpoints (PFS/stable disease and TTE) as their secondary endpoints (Figure 2).

Figure 2.

Clinical endpoints for active and ongoing phase II/III trials since 2018 (a) Primary endpoints (b) All endpoints.

Only ongoing phase II and III trials studying IPF that started on or after January 1, 2018 were included in this figure. Trials with more than one endpoint were accounted for in multiple categories.

Due to the heterogeneous and complex nature of IPF that involves a network of tissues, there is not a single, validated biomarker associated with IPF disease progression or prediction of the disease course. ³³ One of the emerging measurements proposed as a biomarker/surrogate endpoint for IPF is functional respiratory imaging (FRI) of lung and airway structural and functional parameters through HRCT scans. An increasing number of IPF clinical trials have been collecting HRCT scans by which to evaluate their potential as a biomarker for IPF. Notably, for HRCT to become a validated biomarker/surrogate endpoint, correlation to clinical outcomes, such as exacerbations and mortality, is required. In addition, HRCT is typically only measured at baseline and at the end of a study, limiting its ability to be evaluated throughout IPF disease progression over time.^{20,22–24,34} A summary of endpoints being evaluated by ongoing phase II and phase III studies registered on clinicaltrials.gov is illustrated in Figure 2.

Clinical pharmacology considerations

Dosing strategies

Given the challenges and gaps in clinical trial design, it is no surprise that identification of the FIH dose and definition of a proper dose range for patients with IPF are challenges. A pharmacologically active dose in animals does not directly extrapolate to humans, as there is no dose–response, exposure–response, or target engagement that can be directly translated from animal model(s) to human patients. Efficacy endpoints in animals often involve intensive and invasive sampling of tissue that cannot be safely assessed in patients, and as aforementioned, disease severity and manifestation are not identical between species. Despite ongoing efforts, a reliable and validated PD biomarker that can reflect therapeutic changes to disease has not yet been identified. Thus, drugs currently approved and under late-phase clinical trials use surrogate lung function tests as primary endpoints. Neither nintedanib nor pirfenidone trial data demonstrated direct dose–/exposure–responses at the time of drug approval, but instead cited exposure-pulmonary function test correlations as evidence of therapeutic effect.^35–37 While these surrogate endpoints were viewed as evidence of therapeutic efficacy, the correlation between their manifestation and fibrosis severity was not established, and high variability within the patient population was reported.^26,38 As such, the design of a study, inclusion and exclusion criteria, and patient selection at baseline play significant roles in determining whether the study result is statistically meaningful. Furthermore, to observe a meaningful change in these indirect pulmonary function improvements between control and active arms, lengthy trials of 24–52 weeks have been the historical standard, but new studies have started to explore shorter study durations.^12,39 The time and cost of lengthy trials to establish proof of activity may also discourage extensive dose ranging and exploration in the interest of accelerating to pivotal phase and drug approval.

Another challenge in dose justification in IPF is the common practice of indication expansion for anti-fibrotic therapeutics, where the initial FIH/patient dose exploration was not done in patients with IPF. Yet, the safe and pharmacologically active dose identified in other populations may be directly used in IPF studies without further dose exploration. Tissue disposition and exposure to each anti-fibrotic therapeutic can vary significantly, and the optimal dose in one type of fibrosis may not apply to another. Furthermore, while pharmacokinetics between healthy subjects and patients with IPF have generally been consistent, there is limited knowledge in validating that disease does not alter the exposure of drug at the site of action which may be molecule- and mechanism-specific.^40,41 Furthermore, cautious considerations should be taken around the extent of dose extrapolation whenever possible, including evaluation of drug-specific tissue disposition to target organ, the mechanism of action of fibrosis at target organ, and application of suitable preclinical model to further strengthen dose justification.

With limited knowledge of disease progression and a lack of biomarkers, dose selection and optimization are challenges in this field with plenty of opportunities. Advancements in novel approaches have emerged in recent years that have the potential to address challenges of dose extrapolation and optimization. Mathematical models and quantitative systems pharmacology (QSP) models can be leveraged to bridge preclinical and clinical target interaction with investigational drugs for FIH dose justification. ⁴² In this regard, artificial intelligence digital twin platforms combined with a comprehensive disease progression model can significantly accelerate the dose selection and optimization process, as it will allow us to predict how a patient is likely to progress on placebo and on treatment with a smaller trial design. This approach is especially beneficial in trials that study rare diseases, such as IPF.

Modeling and simulation

While several attempts have been made to characterize both disease severity and progression of IPF, the rarity of IPF and the lack of long-term follow-up data make these challenging. Registrational IPF trials commonly finish at around 48–72 weeks from the first dose, despite IPF being a slow-progressing disease.^20,43–46 Age, smoking history, and lung function measurements, such as FVC and %DLCO at baseline, were identified as significant prognostic factors for patients with IPF.^25,47,48

Using the largest pool of placebo-treated subjects from the six registrational IPF trials for pirfenidone (ASCEND, CAPACITY) and nintedanib (TOMORROW, INPULSIS-1, and INPULSIS-2), which included 1132 subjects, a multivariate Cox proportional hazards model revealed that IPF patients with an absolute decline from baseline in % predicted FVC at any time during the study of 15% or more had a sixfold higher risk of mortality than patients with a decline of 5% or less. In addition, patients who experienced at least one IPF exacerbation event at any time during the study had a 10-fold increased risk of death. ²⁵ In another study with a multivariate Cox proportional hazards model using pooled data from all randomized patients with IPFs treated with interferonγ-1b [INSPIRE phase III clinical trial (N = 826)], baseline 6MWD and longitudinal change in 6MWD were also shown to be strongly associated with mortality risks in patients with IPF. ²⁷ Although these models showed statistically significant associations between common IPF clinical trial endpoints, such as % predicted FVC and 6MDW with survival, they did not describe the natural progression or severity of lung fibrosis over time.

One of the first attempts to characterize the disease progression of IPF using % predicted FVC as a surrogate measurement was published in 2021 by Bi et al. ⁴⁹ A nonlinear parametric model with Weibull function was shown to appropriately describe the observed % predicted FVC over time in the placebo and treatment arms. This model pooled data from 2823 patients with IPF across six registrational studies of pirfenidone and nintedanib [TOMORROW (study 30), INPULSIS-1 (study 32), and INPULSIS-2 (study 34) for nintedanib; ASCEND (study 16), CAPACITY (study 4 and study 6) for pirfenidone]. The model did not differentiate which drug the patients received, and it was assumed that both drugs worked by improving the symptoms and modifying the disease time course. The model was able to describe the data in natural disease state (placebo) versus treatment. Notably, this study took into consideration study effects and estimated several parameters specifically for each study due to the heterogeneity of the patient population at baseline, making it difficult to apply for other trials without re-estimating the parameters and limiting its application for prediction beyond the observed range of data.

Given the rarity of IPF and the strict inclusion/exclusion criteria in IPF clinical trials (% predicted FVC ⩾ 50–55%, DLCO > 30–35%),^25,50 little is known about the patients with IPF with mild physiological impairment (defined as patients with IPFs with % predicted FVC ⩾ 80%) and disease progression. Kolb et al. ⁵¹ performed a subgroup analysis of FVC decline rate in patients with % predicted FVC > 90% versus ⩽ 90% and found no significant difference. To this end, IPF data from the Australian IPF Registry (AIPFR) were analyzed to evaluate prognosis markers associated with disease outcome using univariable and multivariable Cox proportional hazards models. Interestingly, 216/416 patients in this registry classified as ‘mild’ had a similar rate of disease progression to the moderate–severe patients despite having better overall survival. Within the mild impairment patient cohort, the only significant prognostic factor which predicted either death or progression at 12 months identified using univariable logistic regression was oxygen desaturation during the 6MWD test; however, the limitation lies in the non-standardized practice of conducting 6MWD test among clinical sites. ⁵² It is also worth noting that less is known about anti-fibrotic efficacy, safety, and disease progression in patients with advanced IPF, as most trials exclude patients below 45% FVCpp and below 30% DLCOpp at screening. A recent real-world data study retrospectively evaluated the patient outcomes between mild–moderate IPF versus advanced IPF patients and found no greater lung function decline over time, but notably higher mortality with advanced disease, and echoed the need to have greater inclusivity in future trials to evaluate anti-fibrotic therapy in patients at all levels of disease progression. ⁵³

A mathematical model of renal interstitial fibrosis developed by Hao et al. ⁵⁴ was later adapted for IPF by including two features unique to IPF: involvement of two distinct populations of macrophages [monocyte-derived inflammatory macrophages (M1) and anti-inflammatory alveolar macrophages (M2)] and the complex geometry of the lung which comprised a large number of alveoli. ⁵⁴ This model suggested the path toward designing a more comprehensive study of IPF, such as evaluation of genetic and epigenetic factors, and combination of multiple drugs targeting different pathways for a more effective treatment of the disease.^55–58

In all the models described above, extensive covariates such as baseline disease characteristics and demographics were evaluated for their contributions to the progression of IPF disease. Importantly, there is a need for predictive biomarkers that are validated to be well correlated with clinical endpoints and survival in patients with IPF. A comprehensive model that uses data from clinical trials, real-world databases, biomarkers, and imagining data, such as the HRCT, would significantly contribute to the understanding of the natural history and disease progression of IPF. One of the biggest challenges in clinical trials for IPF is the determination of the efficacious therapeutic doses. Once a disease progression model for IPF is established, it can be used to simulate how the patients would progress naturally versus on treatment. This information can be used to support dose selection and to predict clinical trial outcomes. This is of particular benefit as IPF is considered a rare disease and recruitment into large randomized clinical trials can be lengthy, costly, and potentially infeasible.

Special populations

IPF is an adult-onset disease usually observed in patients who are 50 years old or above, and while IPF is typically not diagnosed in the pediatric population below 18 years of age, ⁸ some disorders with characteristics of a fibrotic disease have been observed in children. In this regard, although rare, pulmonary fibrosis (PF) had been described in pediatric patients as a specific form of children’s interstitial lung disease (chILD). Accurate estimates of the incidence and prevalence of PF in children are difficult to determine. ⁵⁹ In addition, the course of PF progression is highly variable in children since fibrosis may develop with differing severity and at various stages of lung development. ⁶⁰ The development of lung fibrosis is also difficult to detect in children and is rarely reported due to the lack of established imaging criteria and the challenges in performing biopsies and pulmonary function tests (PFTs) in children. PFTs such as FVC and DLCO are commonly measured as clinical endpoints in trials, yet these measurements require a level of compliance that can be challenging to standardize within and across studies involving children. Taken together, these factors contribute to the lack of data for the understanding of PF in children. Therefore, non-invasive biomarkers that can be established to have significant correlation with clinical endpoints of PF will be helpful for drug development for both adults and young children.

In studies that investigated the prevalence of chronic kidney disease (CKD) in patients with IPF, 30% of patients with IPF had CKD ranging from stage 3 to 5. These patients performed significantly worse in lung function tests and exercise capacity (6MWD) compared to IPF patients without CKD.^61,62 This number is higher than the prevalence of CKD in the general population. ⁶³ Some studies have suggested a mechanistic link between IPF and CKD, as both have been shown to be aging diseases.^64,65 Ethnicity differences may also play a role in the effect of CKD in patients with IPF. For example, while the cause of death for Japanese patients with IPF is typically acute disease exacerbation, 16–27% of deaths in IPF patients from the United States and Europe are due to cardiovascular disease.^66,67 Thus, in IPF clinical trials, renal functions and ethnicity should be evaluated as covariates that contribute to disease progression.

The correlation between impairment in liver function and IPF is also not well established. A large retrospective study of Japanese patients revealed that the annual incidence of IPF among patients with hepatitis C virus (HCV) was significantly greater compared to patients with hepatitis B virus (HBV) ⁶⁸ . This correlation of HCV and IPF was first reported in Japanese and Italian patients but was not observed in a study by Irving et al.^69–71 in HCV patients in the United Kingdom. While these conflicting results may be explained by environmental, genetic, or other factors, there has not yet been a study to support this hypothesis. ⁷² It was also reported that patients with chronic hepatitis C infection had significant reduction in CYP3A4 activities compared to healthy volunteers. ⁷² This can potentially impact the pharmacokinetic exposure of IPF treatments, and as a result might affect efficacy and safety. Moreover, the antiviral concomitant medications used to manage and treat HCV may affect the exposure of IPF treatment and potentially impact drug response and disease progression of IPF. Thus, during IPF drug development, potential effects of liver functions cannot be ignored.

Conclusion/expert opinion

Taken together, it is evident that curative anti-fibrotic therapy remains an unmet medical need in IPF. From a clinical pharmacology standpoint, progress is currently hindered by myriad challenges from preclinical to clinical stages. The lack of suitable nonclinical models currently makes translation of mechanism of action, disease progression, pharmacology, efficacy, and FIH dose extrapolation difficult. Clinically, there are different diagnostic criteria worldwide for IPF, compounded with high variability in disease progression and symptom manifestation, resulting in heterogeneous clinical study populations, which in turn, leads to difficulty in establishing dose–/exposure–response relationships. Differences in regional practice in how some pulmonary function assessments are performed (e.g. no standard method for 6MWD) further complicates matters. In addition, the lack of a clear PD response biomarker contributes to the challenge in enabling robust evaluation of PK/PD, PK–efficacy relationships, or early decision-making concerning dose selection during drug development. Nevertheless, with challenge also comes many opportunities the industry can embrace and put forth effort to overcome with novel technologies and methodologies.

Advancement in imaging technology is valuable in not only early detection and diagnosis of IPF but also in differentiating disease progression and mortality outcomes in patients when traditional measurements of lung function, such as FVC, yield marginal differences. For example, serial CT analysis in IPF patients with marginal FVC decline showed that traction bronchiectasis severity might serve as an independent mortality predictor. ⁷³ Although imaging remains an exploratory endpoint in most IPF clinical trials, the potential for its application in IPF research and drug development is encouraging as more longitudinal data are being collected and quantified in IPF clinical trials. Since IPF is still a relatively rare disease with heterogeneous clinical properties, well-controlled clinical trials often have strict inclusion and exclusion criteria to ensure a clearly defined patient population. Therefore, data from only clinical trials might not provide a comprehensive understanding of IPF with different comorbidities. Real-world data collection and access are becoming increasingly important for studying IPF disease.^74–76 Furthermore, with rapid discoveries in the field of genetics and proteomics, greater insights into the link between genetic risk factors and IPF susceptibility, and phenotypes across populations are being revealed.^77–80 The emergence of high-quality longitudinal data from large-scale studies with diverse cohorts, along with the availability of powerful computational tools, is encouraging for a development of curative treatment for IPF patients in the future.

With new field-wide initiatives promoting greater diversity and increased inclusion in clinical drug development, there is also opportunity to consider study designs that can include patients with IPF who would otherwise receive less attention and who may not have access to the studies. This is especially important as external and environmental exposure to pneumoconiosis has already been reported as significant risk factor in IPF disease, which is also more prevalent in developing countries, compared to developed countries, where most clinical studies for IPF drugs have been conducted. ⁸¹ Patients in developing countries with higher exposure risk often have limited access to facilities and technologies that enable proper diagnosis of IPF that would qualify them for a trial, and the economic burden post-diagnosis can be difficult to overcome. ¹¹ Rise of patient-centric digital technologies may enable widened access to studies, allow data collection from a more diverse patient population, and ultimately, enable more robust dose optimization in all patients with IPF worldwide. Another opportunity for diversity and inclusion is investigating the potential impact of gender and sex as a covariate in IPF development and treatment outcomes. Although historical reports have often reported that IPF is more prevalent in men, and that men are found to have worse disease burden and comorbidities, it is difficult to exclude the possibility of potential confounding factors that have led to more extensive research/inclusion of male subjects in clinical studies. ⁸² In fact, a new study has shown that after correcting for risk factors, the effect of sex on risk of IPF is attenuated. ⁸³ Continuous data collection of disease manifestation and progression to evaluate gender effect can lead to more individualized treatments in the highly diverse IPF population.

In this review, we have summarized the current state of art in clinical pharmacology strategies for the development of IPF therapies. While significant hurdles remain in preclinical to clinical translatability and PD biomarker discoveries, emerging tools and data have opened the potential for a more comprehensive and multifaceted approach to enhance our understanding of IPF disease and its progression for optimization of clinical trials. Real-world data sharing is a key component to enabling more robust and accelerated therapy development, especially for a rare disease, such as IPF. In a joint effort to fulfill the unmet needs for a resolutive therapy for patients with IPF, we encourage broader sharing of knowledge and data across the industry, academia, and health authorities, and continued cross-functional collaborations.